Monitoring Concept Study for Aerospace Power Gearbox DrivetrainDr.-Ing. Sebastian Nowoisky, M. Sc. Mateusz Grzeszkowski,M. Sc. Noushin Mokhtari, Dr. Jonathan G. Pelham and Prof. Clemens Gühmann

IntroductionRolls-Royce is pioneering the UltraFan (UltraFan is a registered trade mark owned by RR PLC) engine family ar-chitecture containing a PGB in a power range of 15 to 80 megawatts (Ref. 12).

The new engine architecture must be more economical than the existing Trent engine family. The consumption of fuel is around 25 percent less than on a Trent 700 power plant (Ref. 13). To increase the efficiency of the Ultra-

Fan, a planetary gearbox is introduced between fan and intermediate pres-sure compressor. This enables running the turbine to rotate faster and allows a reduced fan speed. Subsequently, the bypass ratio will be increased and the

Figure 1 PGB monitoring method.

This paper was first presented at International Conference on Gears 2019, 3rd International Conference on High Performance Plastic Gears 2019, 3rd International Conference on Gear Production 2019, Garching/Munich (VDI-Berichte 2355, 2019, VDI Verlag GmbH, Page 269–286)

46 Power Transmission Engineering ]————WWW.POWERTRANSMISSION.COMAUGUST 2020

TECHNICAL

emitted noise level decreased. To build up monitoring capability for the novel gear technology, it is essential to enable additional digital services as a part of the Rolls-Royce digital strategy (Ref. 4).

The PGB is designed as a planetary gearbox; the gear ratio is used to re-duce the fan speed, and will be able to transfer 100,000 hp (Ref. 14). In 2017 in the test facility at Dahlewitz, dur-ing a test run 70,000 hp was transferred by the PGB, setting record power level (Ref. 11). Due to heavy loads and high operation hours, hydrodynamic jour-nal bearings are going to be integrated (Ref. 2). Hydrodynamic journal bearings offer advantages over roller bearings for high rotational speeds, impact loading or heavy oscillations, and vibrations.

Using a gearbox in a turbojet engine introduces additional failure modes such as gear wear, pitting, and gear teeth cracks (Ref. 1). The rotor’s syn-chronous resampling method is the state of the art for gearbox vibration monitoring and serves to significantly reduce noise. A more advanced wavelet analysis is compared with conventional order analysis or time domain feature extraction. Acoustic emission signals, as well as acceleration signals, are sub-jected to these methods. A trade be-tween methods and signals is made to facilitate the selection of a suitable gear monitoring system.

Power Gear Box (PGB) integrated journal bearings are a focus of this re-search activity. The presented method allows the monitoring of journal bear-ing mixed friction events in a planetary gearbox (Refs. 2–3). The drawback on the previously presented method used with a subscale journal bearing rig ap-plication is that it depends on a mea-surement position close to the bear-ing. Therefore a WDTU is essential to transfer the acoustic emission signal acquired close to the journal bearing across the rotating carrier to the sta-tionary part of the gearbox.

In this paper methods and opportuni-ties will be presented to detect journal bearing mixed friction, as well as how to transfer the data from the rotating to the static part of the gearbox. A mockup was built to support appropriate testing of the WDTU prior to the test on a subscale gearbox being carried out.

Gear Monitoring on PGBAn evaluation of the PGB system design process resulted in the requirement to monitor the gear train to detect pos-sible gear failures such as gear wear, pitting and gear teeth cracks. Depend-ing on the gear failure, several physical measurands — like acoustic emission (AE), acceleration, oil particles, and temperature — can be measured. How-ever, further investigations have shown that AE and acceleration sensors are best suited for early fault detection in the case of the above-mentioned fail-ures, due to their short response time to propagating faults.

Planet-passing monitoring method. Since the vibration signals in the gear train have a vibration periodicity with respect to the gear meshing patterns, the following signal processing method for gear monitoring is shown in Figure 1 using acceleration signals. The accel-eration signals were recorded on the planetary test bench of the FZG of the TU Munich (Ref. 8).

The presented method aims primar-ily at the detection of locally distrib-uted failures on the planet gears and the sun gear. The main idea is to moni-tor the vibrations of the meshing tooth of each planet gear separately to de-tect anomalies in the vibration pat-terns when faulty teeth are in contact. First, the time-dependent vibration signal, which can consist of accelera-tion sensor and AE sensor signals, is merged with the carrier speed sensor and transferred to the angle-domain

using speed-synchronous resampling. This signal processing step reduces the smearing effect in the spectral rep-resentation, which can be caused by speed fluctuations. The low-frequency (LF) vibration signal in the upper-right corner of Figure 1 shows the amplitude modulation effect that results in an in-creased vibration amplitude when a planet gear is directly below the vibra-tion sensor mounted on the ring gear. After the resampling method, the vibra-tion data are windowed with a window width of 360°/Np, where Np is the num-ber of planet gears of the planetary gear. The center of the window is at the carrier angle point if a planet gear intermeshes right near to the vibration sensor. In the preprocessing step, various filters such as low-pass filters, gear mesh band-stop filters for revealing the vibration side-bands, and high-pass filters are applied to the window-shaped vibration signals to feed the feature extraction methods in the last step of Figure 1. Afterwards, features are extracted from the angle-domain vibration signal (e.g.: root mean square (RMS), kurtosis, crest factor) and from the order-domain of the vibration signal (e.g.: FM0, FM4, NB4, NA4, side-band level factor) (Ref. 24).

In addition to the method presented, non-resampled vibration signals are also windowed in the time-domain and used for preprocessing and feature ex-traction to allow a detection of abnor-mal vibration characteristics in the vibration signals that occur through-out the gear train due to gear failures.

Figure 2 Sensor instrumentation for the B2B gear test rig.

47Power Transmission EngineeringAUGUST 2020

Further tests on the FZG planetary gearbox (Ref.8) are planned this year to evaluate the method presented for monitoring the planetary gear train.

Pitting detection. The method pre-sented above was developed to en-able gear failure detection in a plan-etary gearbox. However, since one of the main gear failures, apart from tooth wear and gear teeth cracks, is macro-pitting, a failure-specific investigation must be conducted to provoke pitting. Therefore, the following section pres-ents a pitting detection method that was evaluated with the help of load-carrying capacity tests on a back-to-back (B2B) test bench.

The test rig used is standardized to DIN ISO 14635 (Ref. 20) and consists of a power circuit consisting in particu-lar of test gears, a drive gearbox, a tor-sional shaft and a coupling (Fig. 2). Five different gear variants with respect to their macro geometries, such as pitch deviations and flank roughness, as well as their final manufacturing step, were investigated to demonstrate the ro-bustness of the pitting detection meth-ods (Ref. 1). After a defined running-in phase, the pitting tests were carried out with a constant load of 950 Nm and an engine speed of 2400 min–1.

Since the formation of pitting on the tooth flank leads to a disturbed vi-bration characteristic with periodic loading of the spur gears, vibration sensors were used (Fig. 2). Two iden-tical piezoelectric acceleration sen-sors were placed on the bearing shell of the test gears to measure the radial

accelerations resulting from locally distributed vibration. In addition, one piezo ceramic acoustic emission sensor (AE) was mounted on the bearing shell.

The AE sensor is able to measure high-frequency AE pulses above 50 kHz that are generated by the excitation of elastic waves when a pitted tooth flank is in tooth engagement. In addition, the oil temperature was measured with a Pt1000 thermocouple and the instan-taneous shaft rotation angle of the tor-sional shaft with an incremental mag-netic encoder.

After the measurement data acqui-sition, quantifiable indicators for pit-ting detection are obtained from the data. First, the vibration data is merged with the incremental encoder data and resampled. In the next step, sensor-dependent features are extracted from the acquired sensor signals. The fea-tures are extracted from the four cen-tral moments and the root mean square (RMS) values of the time domain sig-nal and from the spectral power den-sity of the vibration sidebands between the harmonics of the gear mesh order. In addition, significant features are ex-tracted from the coefficients of contin-uous wavelet transform (CWT) and are described in more detail in (Ref. 1). To evaluate the relationship between the pitting progress and the feature val-ues, the Fisher discriminant criterion (Ref. 21) was used. With this criterion value, the features with the highest sen-sitivity to pitting damage can be found.

The kurtosis feature of the CWT coeffi-cients generated by the AE sensor and the

calculated spectral power densities of the vibration sidebands between the 7th and 8th harmonic of the gear mesh order gen-erated by the acceleration sensors provided the most robust features for all gear variants tested. Figure 3 shows the normalized fea-ture values with the highest sensitivity to pitting damage of a test gear variant with a comparatively high single pitch deviation. The complete test results for the other test gear variants can be found in (Ref. 1).

The pitting tests revealed that the fea-tures calculated from acceleration sen-sor data should be obtained from the sideband spectrum, and the features calculated from AE sensor data should be obtained from the CWT coefficients. Although the results did not show that AE sensors allow earlier pitting detec-tion than acceleration sensors, the AE technology has the advantage that the high-frequency range of the AE sen-sor improves the separability between pitting-induced acoustic events in the high-frequency range and test bench vibrations in the low-frequency range. However, due to the higher sampling rate required and the associated higher hardware costs, first of all it must be ex-amined whether the benefit justifies the higher effort.

Journal Bearing MonitoringFor hydrodynamic journal bearings, wear is the most essential damaging mechanism as a result of mixed or dry friction. These friction states are caused by conditions like low speeds, overload, start/stop cycles, insufficient oil supply or oil contamination. A breakdown of this component could have a negative impact on the product reliability, which causes high maintenance costs and downtime. Therefore, the journal bear-ing should be monitored sufficiently.

The literature provides several pos-sibilities to detect damaging friction states of journal bearings, such as mon-itoring the bearing temperature, fric-tion torque, electrical transfer resis-tance between shaft and bearing, oil and vibration analysis (Refs. 17–19). All these possibilities are either not fea-sible for the PGB application or reach their limits for early fault detection. A suitable opportunity to detect mixed or dry friction at an early stage is the use of acoustic emission (AE) technology. The

Figure 3 Pitting detection features.

48 Power Transmission Engineering ]————WWW.POWERTRANSMISSION.COMAUGUST 2020

TECHNICAL

acoustic emission technology provides a sensitive and robust method to detect friction conditions. In this work investi-gations were done at the classification of the three basic friction conditions fluid, mixed, and dry friction based on AE sig-nals and pattern recognition techniques to provide a condition-based mainte-nance for hydrodynamic journal bear-ings. The suitability of these methods for monitoring the degradation state of hydrodynamic journal bearings is also

shown. For this purpose, tests on sub-scale test rigs were done.

To differentiate between the three basic friction states, several speed ramps have been run from high speeds to low speeds at a constant radial load. Figure 4 shows the raw AE signals (top), where the left figure illustrates the fluid friction state and the right figure the mixed friction state. Some modu-lation effects can be seen within one cycle for the mixed friction state. These

modulation effects are indications for mechanical friction.

In (Ref. 10) it is shown that these ef-fects could be used to localize friction within the journal bearing. The continu-ous wavelet transformation (CWT) was used to analyze the acquired AE signals (bottom figures). During mixed friction conditions, besides the low frequencies which occur in fluid friction, also fre-quencies above 100 kHz occur. It can be concluded that in frequency bands lower

Figure 4 Comparison of raw AE signal and CWT of AE signal from fluid friction state (left) with raw AE signal and CWT of AE signal from mixed friction state (right).

Figure 5 RMS and kurtosis of pre-processed AE patterns for a speed ramp at a constant load.

49Power Transmission EngineeringAUGUST 2020

than 100 kHz the differentiation between fluid, mixed, and dry friction is not pos-sible with the AE technology (Ref. 3).

Due to this result, only frequency bands over 100 kHz should be consid-ered. Therefore methods for extract-ing distinct frequency bands (e.g. CWT, STFT, etc.) should be used as pre-pro-cessing steps before extracting features for the classifier. Several features were extracted from the pre-processed pat-terns. The most promising features are the RMS and the kurtosis for differenti-ating between the friction states (Fig. 5). These results and methods were veri-fied at other subscale test rigs. The ex-tracted features can now be used as the input for classifiers. The support vec-tor machine classifier offered good re-sults for the classification of the friction states.

To monitor a journal bearing it is not enough to only identify the actual fric-tion state; the actual degradation state of journal bearings should be analyzed as well. The degradation of journal bearings is typically characterized by the actual wear depth. Short time wear experiments were done at the TU Berlin test rig. During these tests the journal bearing was driven for an overall time of 18 hours at constant operating con-ditions to generate wear.

After every step of 2 hours, identical speed ramps were driven and the pat-tern recognition methods were applied. Figure 6 shows that the feature changes over the degradation state. This result shows that it is possible to monitor the degradation state, which is the actual wear depth, by using the AE technology

in combination with suitable pattern recognition techniques (Ref. 2).

In order to be able to predict the re-maining useful lifetime (RUL) of jour-nal bearings, long-term wear tests must be done. These experiments are part of current research work. For this pur-pose, the TU Berlin test rig has been modified so that long-term experiment conditions can be set.

One specific challenge when using the AE technology for the PGB appli-cation is the position of the AE sensor. The nearest static part of the PGB is the ring gear on which the sensor can be mounted. However this position is not acceptable because the gear mesh con-tact between ring gear and planet gears acts as a low-pass filter so that high-fre-quency friction signals are filtered out. In order to successfully apply the devel-oped monitoring methods on the PGB, the AE sensor must be mounted as close as possible to the mixed-friction area; this area is rotating, which makes a wireless data transfer necessary.

Wireless Data Transfer UnitAs depicted on the left-hand side in Fig-ure 7 — the generic arrangement of a planetary-style gearbox with the bear-ing between planet and carrier.

As stated previously, an acoustic emission sensor must be able to mea-sure up to 1 MHz bandwidth (25) in a position close to the journal bearing. Hence a signal transfer from the rotat-ing carrier to the static part of the gear-box is required. Beginning with com-mon environmental requirements of a gearbox, the data transfer must deal with an oily environment, tempera-tures above 120°C, and strong vibra-tions caused by the gear train. A system engineering-based assessment was carried out to understand the require-ments of a WDTU solution. As a result, a sensor node containing comprehen-sive electronics, e.g. — as presented in (Ref. 16) — can’t be applied. Currently available high-temperature electron-ics either do not meet the functional re-quirements or the expected reliability

Figure 6 Degradation levels indicated by AE RMS for speed sweeps from 40 to 400 rpm, constant load of 1,500 N (left). Average surface roughness over different speed steps (right).

Figure 7 Schematic of planetary gearbox and applied wireless data transfer unit (WDTU).

50 Power Transmission Engineering ]————WWW.POWERTRANSMISSION.COMAUGUST 2020

TECHNICAL

levels. However, in further investiga-tions a WDTU could be identified that had already been tested on PUMA heli-copter main gearboxes (Refs. 22, 6 or 15) .The WDTU from Cranfield University [UK] meets the requirements of the measurement chain for the gearbox en-vironmental conditions.

The WDTU works on a principle sim-ilar to homodyne radar and, in some respects, to a near-field RFID (radio frequency identity device) (Ref. 23); the system is depicted in Figure 8. The transceiver (A) generates the carrier signal and transmits it through a coax-ial cable (1) to the matching network (B) that is used to tune the character-istic impedance of the static antenna with which it is connected by a pair of wires (3). The static coil transmits this carrier wave to the rotating antenna, where it is picked up by a twisted pair of wires (4) and rectified and smoothed within the processing unit (C). This creates a DC voltage which is used to power an amplifier and filter stage that is connected by a cable (5) to the sensor (D). This is where it differs from RFID, as this sensor signal — after it has been filtered and amplified — is used to tune the resonant frequency of the rotating antenna using a varactor diode. This al-lows a continuous signal to be transmit-ted that is representative of the desired signal from the sensor.

The static antenna detects this reso-nant frequency modulation as back-scatter of the transmitted signal, which is then sent back to the transceiver (A) via a coaxial cable (2) where it can be recovered by comparison to the car-rier wave signal used to drive the static antenna.

The design of the system presented some challenges due to the mass and size constraints combined with the environment within which the device needs to operate. Additional complex-ity was created by the need to include good RF shielding and grounding when sensor elements can create a ground loop through the device under test.

The combination of journal bearing mixed friction detection and WDTU enables a detailed monitoring of a JB inside a planetary gearbox. A mockup test was executed to represent the inte-gration of the antennas in the subscale

planetary gearbox to demonstrate the ability of the WDTU to cope with the speeds and perform the signal trans-fer. The dominant gear mesh signal is represented by a triangular wave signal and an estimated 1,000 times smaller sinusoidal signal at 277.85 kHz repre-sents occurring mixed friction of the journal bearing.

The test demonstrates that the mixed friction signal can be transferred and extracted (Fig. 9). The red dotted lines highlight at 5 kHz the gear mesh from the signal generator. The signal from the simulated mixed friction can also be seen highlighted by the dotted green line at 277.85 kHz. What remains, is the ratio of the dominant gearbox gear mesh and the subsequent harmonics compared to the mixed friction pattern.

Conclusion and OutlookIn this work new monitoring concept studies are presented for a power gear-box integrated into a TurboFan power

plant (UltraFan). Different methods were assessed to monitor gear vibration and the associated benefits and draw-backs have been identified. For dynamic engines such as a jet engine, rotor syn-chronous resampling is recommended due to rotational speed fluctuations and the resulting smeared frequency spec-trum. This helps to significantly mini-mize the signal noise and to detect an anomaly in the order spectrum earlier than without resampling.

Acoustic emission sensors provide a sensitive method to monitor the jour-nal bearing for different purposes such as mixed friction detection, coating degradation, and coating defects. In a planetary gearbox a wireless data trans-fer unit is required to measure close to the bearing friction area up to 1 MHz bandwidth.

A proven technical data transfer so-lution could be identified, which had been demonstrated on helicopter ap-plications with roller bearings. The

Figure 9 Test result sub-scale WDTU at 2200 min-1.

Figure 8 WDTU system overview.

51Power Transmission EngineeringAUGUST 2020

WDTU is modified to accommodate the requirements of the power gear-box and the purpose to measure mixed friction conditions in journal bearings. Further tests are planned on a sub-scale planetary gearbox (Ref. 8) to verify, that the presented methods are capable to achieve monitoring capability for gears and journal bearings.

Acknowledgment. Thanks to Julian Worskett of JWD Ltd. for his invalu-able radio communications expertise. Thanks to the team from the FZG of Dr. Michael Otto and Uwe Weinberger for the support to integrate the WDTU into the sub-scaled gearbox and the test op-portunity. Thanks to the team of WZL — Dr. Christoph Löpenhaus and Philipp Scholzen for the support during the pit-ting test campaigns. Many thanks to the participating colleagues and research-ers involved in this monitoring study work for a planetary gearbox. We also acknowledge financial support through funding from the BMWi within the Luftfahrtforschungsprogramm LuFo.

For more information.Questions or comments regarding this paper? Contact Sebastian Nowoisky at Sebastian.Nowoisky@Rolls-Royce.com.

References1. Grzeszkowski, M., P.Scholzen, S. Nowoisky,

C. Gühmann, C. Löpenhaus and G. Kappmeier. “Experimental Study on the Pitting Detection Capabilities for Spur Gears Using Acoustic Emission and Vibration Methods,” in AGMA Fall Technical Meeting, Chicago, 2018.

2. Mokhtari, N., C. Gühman and S. Nowoisky, “Approach for the Degradation of Hydrodynamic Journal Bearings Based on Acoustic Emission Feature Change,” in Prognostic and Health Management, Seattle, 2018.

3. Mokhtari, N., F. Rahbar and C. Gühmann. “Differentiation of Journal Bearing Friction States and Friction Intensities Based on Feature Extraction Methods Applied on Acoustic Emission Signals,” Technisches Messen, Vol. 84, no. S1, p. 6, 2017.

4. Norris, G. “Aviationweek,” 06 02 2018. [Online]. Available: http://aviationweek.com/commercial-aviation/rolls-unveils-intelligentengine-digital-strategy. [Accessed 08 10 2018].

5. Mokhtari, N., M. Grzeszkowski and C. Gühmann. “Vibration Signal Analysis for the Lifetime Prediction and Failure Detection of Future Turbofans Components,” in Schwingungen in Rotierenden Maschinen (SIRM), Graz, 2017.

6. Fang, D., M. Greaves and D. Mba. “Helicopter Main Gearbox Bearing Defect Identification with Acoustic Emission Techniques,” in IEEE Prognostics and Health Management, Ottawa, 2016.

7. Greaves, M. “Towards the Next Generation of HUMS Sensor,” in ISASI, Adelaide, 2014.

8. TUM. “Ein Getriebe für 100.000 PS,” 26 04 2018. [Online]. Available: https://www.youtube.com/watch?v=Zhyz3A6deZ4. [Accessed 11 10 2018].

9. Grzeszkowski, M., C. Ende and S. Nowoisky. “Verfahren und Vorrichtung zur Überwachung der Kinematik eines Epizyklischen Planetengetriebes,” Deutschland Patent DE102017121239.6, 2017.

10. Mokhtari, N. and S. Nowoisky. “Verfahren und Vorrichtung zur Überwachung eines Gleitlagers,” Deutschland Patent DE 10 2017 119 543, 2019.

11. Norris, G. “Aviationweek,” 05 09 2017. [Online]. Available: https://aviationweek.com/commercial-aviation/rolls-gearbox-hits-record-power-level. [Accessed 05 04 2019].

12. Norris, G. Aviationweek, 27 05 2015. [Online]. Available: https://aviationweek.com/technology/rolls-freezes-design-first-ultrafan-test-gear. [Accessed 05 04 2019].

13. Ebner, U. “Flugrevue,” 12.10.2017, [online], Available: https://www.flugrevue.de/flugzeugbau/prototyp-im-test-rolls-royce-ultrafan-getriebe-gigant/. [Accessed 05 04 2019].

14. Bennet, J. Popular Mechanics, 28.10.2016, [online], Available: https://www.popularmechanics.com/flight/a23603/rolls-royce-tests-poweful-turbofan-gearbox/. [Accessed 05 04 2019].

15. Mba, D., M. Greaves, L.Tauszig and E. Faris. “Novel Internal Sensors for Helicopter Main Rotor Gearboxes,” 40th European Rotorcraft Forum, at Southampton, UK, September 2014.

16. Funck, J., R. Knoblich, D. Scholz, S. Nowoisky and C. Gühmann. “Drahtloses Messsystem für Temperaturen in Kraftfahrzeugkupplungen,” TM – Technisches Messen, 2013.

17. Deckers, J. “Entwicklung einer Low-Cost Körperschallsensorik zur Überwachung des Verschleißverhaltens von Wälz-oder Gleitgelagerten Kreiselpumpen Kleiner Leistung,” Ph.D. thesis, Gerhard-Mercator-Universität Duisburg, 2001.

18. Raharjo, P. “An Investigation of Surface Vibration, Airborne Sound and Acoustic Emission Characteristics of a Journal Bearing for Early Fault Detection and Diagnosis,” Ph.D. thesis, University of Huddersfield, May 2013.

19. Fritz M., A. Burger and A. Albers. Schadensfrüherkennung an Geschmierten Gleitkontakten Mittels Schallemissionsanalyse,” Institut für Maschinenkonstruktionslehre und Kraftfahrzeugbau, Universität Karlsruhe, 2001.

20. DIN ISO 14635 Teil 1; Mai 2006. Zahnräder. FZG-Prüfverfahren. FZG-Prüfverfahren A/8, 3/90 zur Bestimmung der Relativen Fresstragfähigkeit von Schmierölen.

21. Bishop, Christopher M. (2009), Pattern Recognition and Machine Learning, Corr. at 8. print. New York, NY: Springer (Information science and statistics).

22. Greaves, M., F. Elasha and et al. “EASA Final Report: Vibration Health or Alternative Monitoring Technologies for Helicopters,” 04.06.2015, EASA_REP_RESEA_2012_6.

23. Want, R. “An Introduction to RFID Technology,” Pervasive Computing, January-March 2006, pp 25–33, DOI: 10.1109/MPRV.2006.2.

24. Vecer, P., M. Kreidl and R. Smid. (2005): “Condition Indicators for Gearbox Condition Monitoring Systems,” In: Acta Polytechnica Vol. 45 No. 6/2005, p. 1–6.

25. M. Greaves, F. Elasha, J. Workett, D. Mba, H. Rashid and R. Keong, “Vibration Health

or Alternative Monitoring Technologies for Helicopters,” European Aviation Safety Agency, Cologne, 2012. (www.easa.europa.eu/sites/default/files/dfu/EASA_REP_RESEA_2012_6.pdf)

Dr.-Ing. Sebastian Nowoisky studied electrical engineering focused on measurement and automation technologies at the TU Berlin (TUB) and the University of applied science in Brandenburg (FHB). He received a diploma in 2008 at the TUB (Dipl.-Ing.) and in 2006 at the FHB (Dipl.-Ing. (FH)). He is working at Rolls-Royce Deutschland Ltd & Co KG since 2014 as System Architect and is the tech lead coordinating research activities to develop a gearbox diagnostic system. In 2016 he became a Ph.D. at the TU Berlin in the field of mechatronic parameter identification of gearboxes. He previously worked 5 years at the TU Berlin as research assistant focused on the field of automated car gearboxes and building up a gearbox test environment. In his professional experience, since 2008 Nowoisky has worked as project engineer in the maintenance department at ThyssenKrupp Metal Forming.

M. Sc. Mateusz Grzeszkowski is currently a research assistant at the Chair of Electronic Measurement and Diagnostic Technology of the Technical University of Berlin with the emphasis on pattern recognition methods for the monitoring of spur gears and planetary gears. He has received a Bachelor and Master Degree at TU Berlin in the field of Electrical Engineering. His experience in machine monitoring started during his master thesis, where he developed a diagnosis system for monitoring a rail wheelset axle using a non-destructive acoustic emission measurement technology at Bundesanstalt für Materialforschung und-prüfung (BAM). After graduating, Grzeszkowski started his research work with the topic on diagnostic methods for monitoring planetary gears at the Technical University of Berlin in cooperation with Rolls-Royce Germany.

M. Sc. Noushin Mokhtari studied mechanical engineering at Leibniz University of Hanover and aerospace engineering at the Technical University of Braunschweig. Since 2015 she works as a research assistant at the Chair Electronic Measurement and Diagnostic Technology of Technische Universität Berlin. Mokhtari’s areas of expertise include the data-based diagnosis and remaining useful lifetime prediction of hydrodynamic journal bearings, as well as signal processing, pattern recognition and classification of acoustic emission signals.

52 Power Transmission Engineering ]————WWW.POWERTRANSMISSION.COMAUGUST 2020

TECHNICAL

Dr. Jonathan G. Pelham is a Research/Teaching Fellow in Flight Data at the Cranfield University Safety and Accident Investigation Centre. In his role as a CSAIC Research/Teaching Fellow in Flight Data, he researches the use of machine learning and data analytics to better understand aircraft flight and improve methods to learn from operations, as well as how to investigate incidents and accidents using UAS (Unmanned Aerial Systems). He has a Ph.D. from Cranfield University, a BEng in Aerospace from Sheffield University and a Postgraduate Certificate in Robotics from the University of Western England. His Ph.D. research focus was the application of IVHM (Integrated Vehicle health Management) techniques to UAS using artificial immune systems to identify faults throughout the aircraft’s life. This was a large-scale drone mission simulation within MATLAB with 10000+ hours of simulated flight including multiple scenarios and mishap modes drawn from accident reports of in-service unmanned aircraft. Pelham has a wide variety of industrial experience from sources as diverse as Mclaren Racing and the UK Queen Elizabeth Class carrier project, where he was the Lead Engineer for Fire Detection systems onboard ship.

Prof. Dr.-Ing. Clemens Gühmann studied electrical engineering and has been head of the Chair Electronic Measurement and Diagnostic Technology since 2003 at the Technische Universität Berlin. Gühmann’s areas of expertise include the measurement technology and data processing as well as the diagnosis, predictive maintenance, modeling, simulation and automatic control of mechatronic systems.

In last month’s issue:• Condition Monitoring for Gear

Drives and Gearmotors• Motors for Cooling Towers• Dynamic Behavior of

Planetary Gears• Hybrid Step Motors

… and more!

MISSINGA PIECE?We’ve got you covered! Go to powertransmission.com to see what you missed in last month’s issue, plus another thirteen years of back issues, industry and product news, and more!

aerospace gearboxes

For Related Articles Search

at www.powertransmission.com

53Power Transmission EngineeringAUGUST 2020